Characterize RF Noise Components Using Equivalent Noise Temperature

Previously, we discussed that the noise figure is the commonly used noise specification in RF work. An alternative way of characterizing the noise performance of RF components and systems is the equivalent noise temperature, which will be the main focus of this article.

In general, the noise figure and equivalent noise temperature both provide the same information; however, you might be a little less familiar with the concept of noise temperature. The noise temperature is mainly used in non-terrestrial applications, such as radio astronomy and space-oriented radio links that deal with very small noise levels. Despite its niche application, familiarizing ourselves with the noise temperature concept can give us a clearer picture of how noise figure measurement instruments actually work. In fact, an automatic noise figure analyzer might perform many of its internal computations in terms of noise temperature.

We can use the noise temperature concept to specify the noise produced by a one-port device, such as an antenna or a noise source. To better understand this, consider an arbitrary white noise source with an output impedance of R connected to a matched load resistor, RL, as shown in Figure 1(a) below.

Let's assume that the noise source delivers a noise power of No to RL = R (i.e., the maximum available noise power of the noise source is No). We know that the available noise power of a resistor is kTB. Equating kTB with No, we can find the temperature where the resistor exhibits an available noise power of No.

\[T_e = \frac{N_o}{kB}\]

This observation gives us the noise model shown in Figure 1(b), where a single resistor, R, at a temperature of Te is used to produce the same amount of noise as the original noise source, where Te is the equivalent noise temperature of the noise source. Note that the noise temperature doesn't indicate the physical temperature of the resistor, as you would measure it with a thermometer. The noise temperature is just a concept that allows us to model the actual noise level that a component produces. It's also worthwhile to mention that, by definition, the noise figure concept cannot be applied to one-port devices.

The noise temperature concept can also be used to describe the noise performance of a two-port network. As an example, consider a noisy amplifier with a gain, G, and a bandwidth, B, connected to a matched source resistor, as shown in Figure 2(a).

Next, the available noise at the output of the amplifier can be described using Equation 1.

\[N_o = N_{o(added)} + kT_0BG\]

Where:

Similar to the one-port example, we wish to model the noise from the amplifier by finding a new temperature for the source resistor. To this end, we first find the input-referred noise of the amplifier:

\[N_{i(added)}=\frac{N_{o(added)}}{G}\]

Equating the above value with kTeB gives us the equivalent temperature where the available noise power of a resistor is equal to the input-referred noise of the amplifier in Equation 2.

\[T_e=\frac{N_{o(added)}}{kBG}\]

From this, we can assume that the amplifier is noiseless and, instead, increase the initial temperature of Rs by Te to account for the amplifier's noise. This is illustrated in Figure 2(b).

Now, let's verify our model by calculating the total output noise. Referring to Figure 2(b), we have:

\[\begin{eqnarray}N_o &=& kTBG = k(T_0+T_e)BG \\&=& k(T_0+\frac{N_{o(added)}}{kBG})BG \\&=& kT_0BG +N_{o(added)}\end{eqnarray}\]

Which is consistent with Equation 1 (no big surprise!). Having the noise temperature of the amplifier, Te, we are able to find the noise temperature of the whole system, including both the source impedance, Rs, and the amplifier given by T0 + Te. Additionally, by combining Equation 2 with the noise factor definition below, we can obtain a useful equation that expresses the noise figure in terms of the equivalent noise temperature, shown in Equation 3.

\[\begin{eqnarray}F&=& 1 + \frac{N_{o(added)}}{N_{o(source)}} \\&=& 1 + \frac{kT_eBG}{kT_0BG} \\&=& 1 + \frac{T_e}{T_0}\end{eqnarray}\]

A cascaded system made of N two-port devices is shown below in Figure 3.

Where:

With that in mind, we know that the noise factor of the cascaded system is:

\[F = F_1 + \frac{F_2 - 1}{G_1} + \frac{F_3 - 1}{G_1 G_2} + \dots + \frac{F_N - 1}{G_1 G_2 \dots G_{N-1}}\]

Applying Equation 3, we can replace each Fi term with its equivalent noise temperature and find the noise temperature of the cascaded system:

\[T_{cas} = T_1 + \frac{T_2}{G_1} + \frac{T_3}{G_1 G_2} + \dots + \frac{T_N}{G_1 G_2 \dots G_{N-1}}\]

If Ts denotes the noise temperature of the source impedance, the noise temperature of the whole system—including Rs and the cascade—is Ts + Tcas.

Now, let's look at a few examples to clarify the above concepts.

Assume that for a source temperature of Ts = 60 K, the overall system noise temperature is 380 K. Find the noise figure of the cascade.

The noise temperature of the cascade itself is readily found as Tcas = 380 - Ts = 320 K. Next, we apply Equation 3 to find the required cascade noise figure:

\[\begin{eqnarray}F &=& 1 + \frac{T_e}{T_0} \\&=& 1 + \frac{320}{290}=2.1 =3.22 \text{ }dB\end{eqnarray}\]

Assume that the source noise temperature is Ts = 150 K. Also, suppose that the cascade noise factor, gain, and bandwidth are respectively Fcas = 1.8, G = 6 dB, and B = 10 MHz. Find the available noise power at the output of the cascade.

We first use Equation 3 to find the noise temperature of the cascade:

\[T_{cas} = (F_{cas}-1)T_0=(1.8-1)\times 290=232 \text{ } K\]

Therefore, the noise temperature of the overall system is Tsys = Ts + Tcas = 150 + 232 = 382 K. Finally, we have:

\[\begin{eqnarray}N_o &=& k(T_s + T_{cas})BG \\&=& 1.38 \times 10^{-23} \times 382 \times 10 \times 10^{6} \times 10^{0.6} \\&=& 2.099 \times 10^{-13} \text{ }W = -96.8 \text{ } dBm\end{eqnarray}\]

In a previous article, we looked at the plot of total output noise versus the source resistance temperature, T (Figure 4).

This curve allows us to better understand an important difference between the noise figure and noise temperature. The noise figure metric corresponds to a standard temperature of T0. It actually specifies the ratio of the output noise contributed by RS at T0 (i.e., kT0BG) to that of the device under test, No(added). As can be seen from the figure, this ratio changes with T, and that's why the noise figure is given at a standard temperature. From Equation 2, however, the noise temperature directly specifies the noise added by the device under test No(added), which doesn't change with T. This feature allows us to simply add the noise temperature of a component to the arbitrary noise temperature of the source resistance; and use the overall noise temperature of the system to find the output noise power.

On the other hand, applying the noise figure concept can be a little bit tricky when the source temperature, Ts, is not the same as the standard temperature, T0. If Ts ≠ T0, we cannot directly use the noise figure definition to find the total output noise. In this case, we should first use the noise figure equation to find No(added) and then use that information to find the output noise.

Table 1 gives the noise temperatures for some example noise figure values.

NF (dB)

F

TN (K)

0.5

1.122

35.4

0.6

1.148

43.0

0.7

1.175

50.7

0.8

1.202

58.7

0.9

1.230

66.8

1.0

1.259

75.1

1.1

1.288

83.6

1.2

1.318

92.3

1.5

1.413

120

2.0

1.585

170

2.5

1.778

226

3.0

1.995

289

3.5

2.239

359

Note that for very low noise systems, the noise temperature is a higher-resolution description of the noise performance. For example, while the noise figure changes from 0.5 dB to 1 dB, the noise temperature changes over a relatively larger range, from 35.4 K to 75.1 K. The noise factor also has a small change in this range, going from 1.122 to 1.259. Being a higher-resolution representation, the noise temperature is usually used in characterizing satellite communications systems that deal with exceptionally low noise levels.

As the final section of this article, let's take a brief look at some of the factors that can have an impact on the noise temperature of an antenna. The electrical model of an antenna used as a receiving element is shown below (Figure 5).

The voltage source, VAnt, represents the ability of the antenna to collect signals. RAnt actually models the matching property of the antenna that matches the characteristic impedance of free space to that of our circuitry. The antenna also picks up both signal and noise components impinging on it.

To model the collected noise, we assume that RAnt is at a noise temperature of TAnt. The noise the antenna picks up, and consequently TAnt, depends on several different factors, such as the position of the antenna, its elevation angle, and frequency of interest. For example, if the antenna is oriented toward an electronic device that generates electromagnetic interference (EMI), we expect to collect a greater amount of noise power. However, repositioning the antenna away from the noise source can reduce the noise level.

The antenna elevation off of the horizon is also an important parameter. In ground-to-ground radio links, the antenna is pointed at the horizon. Thus it picks up thermal radiation from the ground, leading to a typical noise temperature of about 290 K, which is the standard temperature used in the definition of noise figure.

On the other hand, in satellite communications, the antenna is pointed at the sky, and the equivalent noise temperature is normally much lower, typically about 50 K. That's why satellite communications systems deal with exceptionally low noise levels and usually characterize their systems using the noise temperature metric. The antenna noise temperature also changes with frequency. Figure 6 shows the noise temperature versus the frequency for an antenna with an elevation angle of 5°.

The noise figure and noise temperature are interchangeable characterizations of the noise performance. The noise temperature concept is mainly used in non-terrestrial applications, such as radio astronomy and space-oriented radio links that deal with very small noise levels. Also, familiarity with the noise temperature concept can give us a clearer picture of how noise figure measurement instruments actually work. In the next article, we’ll discuss one of the commonly used methods for noise figure measurement, namely the Y-factor method, which extensively uses the noise temperature concept.

To see a complete list of my articles, please visit this page.